Bipolar disorder is a common mental illness characterized by one or more periods of mania and depression. Family and adoption studies support the existence of genetic factors in the expression of the disorder in a significant percentage of affected individuals. The ultimate goal of this proposal is to identify a gene or genes that predispose to the development of bipolar disorder. The studies proposed will examine 2 extended pedigrees, one Amish and the other Costa Rican. In the Amish pedigree individuals with bipolar disorder have been clearly identified, and lymphoblast cell lines are available for genetic analysis. Since recent work has put earlier linkage findings into question, we propose to screen the genome for linkage. To minimize problems that might arise from the existence of more than one gene predisposing to BP even in this limited population we will confine our attention to kinships with one affected ancestor. A distinctive feature of our approach is to study only clearly affected members, and to do this intensively using novel somatic cell genetic and molecular techniques to substantially increase the amount of information that we gain from each individual. This will be done by examining individual chromosomes from each affected individual in somatic cell hybrids, and by developing techniques to make studies with all markers informative by detecting polymorphisms specifically present within this particular pedigree. The Costa Rican pedigree has a high prevalence of bipolar disorder and contains over 500 living members. Preliminary studies indicate that 30 of 130 individuals satisfy diagnostic criteria for major affective disorder in a manner consistent with autosomal dominant transmission. We propose a detailed diagnostic assessment of this population, establishment of cell lines from all members, and linkage studies using both the conventional RFLP approach and, when necessary, the novel techniques that we are developing in work with the Amish material.